Attributes & Derivatives¶
This document is part of the Frontiers Design Documentation, a set written to teach instead of to instruct. Each entry takes one design problem, explains what that problem is in the wider language of tabletop design, surveys how games have solved it, and then shows what Frontiers does as one worked answer among many.
It is written for several readers at once:
- Designers using Frontiers who want to understand why it works the way it does.
- Designers using Frontiers who want to change it, and need to know what a change will cost.
- Designers building something else entirely, who want the design space mapped before they choose.
Frontiers is a toolkit, and a toolkit is only useful if you can see what each tool is for. None of this is a rule; the rules live in the Rulesets, and what follows is the reasoning, the history, and the alternatives.
There will be terminology here that is unfamiliar to casual players. Treat this documentation as advanced content, written for designers interested in building. It will not teach you every term or phrase it uses, but it will do its best to define them as they arrive.
What are Attributes¶
Attributes are a game's answer to what a character fundamentally is, before training, tools, or circumstance. Nearly every system has some version of them, and nearly every system derives other numbers from them, which makes this the load-bearing layer of a character sheet: change an Attribute and a dozen other values move.
To give a well-known example, Mongoose Publishing's Traveller gives every character six characteristics, from Strength through Social Standing, each rolled at creation and rated on a small numeric scale. They govern rolls, they gate careers, they take damage in combat, and they decay with age. A Traveller character's whole life is legible from those six numbers, and the game's structure is built to make them matter in more than one way.
What makes Attributes distinct from Skills is that they set the floor while Skills supply the bonus. A character with no relevant training still has an Attribute to fall back on, and that fallback is what a system defines when it picks its list.
Three questions decide the shape:
- How many, and what do they cover? How finely innate capability is divided, and whether every meaningful action has an obvious home.
- What scale do they sit on? Whether the number is used directly, converted into a modifier, or read as a chance.
- What is built on top of them? How much of the character sheet is derived instead of chosen, and how much a single point is therefore worth.
Why Attributes matter¶
Attributes are the most quoted numbers in any system, and the hardest to change later.
The list is a claim about what kinds of characters can exist. A game with separate Strength and Constitution is saying those can meaningfully differ; a game with one Body stat has decided they can't. Every split you make is a build direction you've opened, and every merge is one you've closed.
The list also has a well-known failure mode. Physical Attributes usually find obvious work, while mental ones often govern a narrow slice of rolls and end up as dump stats, the thing nobody invests in because nothing on the sheet depends on it. Which Attribute is weakest is usually a symptom of what the Derivatives forgot to use.
The second half of this document takes that up. Derivatives are where Attributes acquire their real weight, since every value calculated from an Attribute is one more thing a point buys you, and the map of what derives from what decides which Attributes are good.
How much of a character should be chosen, and how much should follow from the choices already made?
The Approaches¶
Attribute systems differ on two decisions made together: how many stats there are, and how much other machinery hangs off them. A game with many stats and heavy derivation feels nothing like one with few stats and none, even when the list looks similar on paper.
A few shapes worth knowing.
Six stats, heavily derived¶
The most familiar shape. Six Attributes feed a large derived sheet, so a single point of a stat moves several downstream numbers at once. Wizards of the Coast's 5e SRD is the reference point most designers share, and a great many games inherit its structure directly.
The appeal here is depth and familiarity. Six stats give room for fine distinctions, derivation makes each point feel consequential, and most players already know how to read the sheet. The cost is the dump stat. With six Attributes, some will always carry more downstream weight than others, and the ones that carry least become the safe place to put a low number. The intelligence-versus-wisdom question, which many systems in this family never resolve convincingly, is a symptom of the same problem: once you have six, telling them all apart is hard.
A wide characteristic spread on a flat scale¶
Some systems carry many characteristics but derive little from them, using the stats mostly as inputs to a skill roll, not as the root of a calculated sheet. Chaosium's Call of Cthulhu works this way: an investigator has a broad set of characteristics, but the game's weight sits in the skill list, and the characteristics mostly feed it.
That keeps a large, textured set of numbers without the derivation cascade, so a point of a stat rarely ripples far. The cost is that some characteristics end up doing almost nothing on their own, present for completeness or flavor rather than because the system leans on them.
Few stats, each doing double duty¶
Modern systems often cut the count and make the survivors work harder. Monte Cook Games' Cypher popularized the version I find most elegant: three stats, Might, Speed, and Intellect, that are each both a rated capability and a pool you spend from to push rolls, absorb harm, and fuel abilities. The stat measures how good you are and doubles as a resource you burn.
Fewer stats leaves nowhere for a dump stat to hide, since three can't conceal a weak one the way six can, and the double duty gives every stat constant relevance. The cost is that each stat now carries a lot of conceptual load, so the lines between them have to be drawn carefully, and a system with this few has less room to express fine distinctions between characters.
Three stats, no derivation¶
The minimalist end. Free League's Into the Odd and its relatives use three stats and derive almost nothing from them: the number is the number, and the rest of the character comes from equipment and fiction.
It's the fastest possible version. Nothing needs calculating, creation takes minutes, and a stat means exactly what it says. The trade is that characters differentiate through gear and play rather than through their sheet, so two characters with identical stats are told apart by what they carry and do.
No Attributes at all¶
A few systems remove the layer altogether. Evil Hat's Fate has no attribute line; a character is defined by skills and aspects, and innate capability is folded into those. There's no "before training" number because the game doesn't model that separation.
The gain is focus, since everything on the sheet becomes something the character does, with no baseline stats sitting behind it. The cost is that the floor Attributes usually provide goes with them. Without an innate number to fall back on, an untrained action has to be handled some other way, usually through the fiction or a default rating baked into the resolution system.
Running under it all is a trade between count and weight. Many stats give you fine distinctions and a rich derived sheet, at the price of some of them being ignorable. Few stats guarantee every one matters, at the price of range. Modern design has broadly trended toward fewer stats that each carry more, and the reason is the dump stat: it is easier to make three or four Attributes all matter than to make six all matter, and a stat nobody invests in is a stat that isn't really there.
Attributes & Derivatives in Frontiers¶
Now, with that overview to catch you up to speed, here's the approach Frontiers takes and the reasoning that led there.
Frontiers uses four Attributes, each read two ways, with most of the sheet derived from them:
| Icon | Attribute | Covers |
|---|---|---|
| ❖ | Vigor | Bodily power and endurance |
| ✩ | Deft | Bodily precision and speed |
| ◉ | Acuity | Mental sharpness and reach |
| ◬ | Resolve | Mental fortitude and steadiness |
Every Attribute has a Score (used without a roll) and a Rank (added to a roll). Full rules live in the Rulesets; what follows is why the shape looks like this.
Four, because every one has a job¶
The common count is six, and six is where the dump stat problem lives. Most six-Attribute systems can't fully justify their own list, and the intelligence-versus-wisdom divide is the usual evidence: two stats that are hard to tell apart, one of which many characters ignore.
Frontiers settled on four for a simple reason. Every Attribute has to do a job no other Attribute does, and every Attribute has to matter to every character regardless of their build. There's no Attribute a martial character can safely dump and none a scholar can, because each one drives something everyone needs. Four is where that held. Earlier development carried six, and the two that got cut were the two that couldn't answer what they did that the others didn't.
The two axes¶
The four sit on a grid, physical against mental and power against precision:
| Power & Endurance | Precision & Reaction | |
|---|---|---|
| Physical | ❖ Vigor | ✩ Deft |
| Mental | ◬ Resolve | ◉ Acuity |
The grid keeps the set complete without overlap, with four quadrants, four Attributes, and neither a gap nor a redundancy. It also sidesteps the stat every system struggles with. Charisma, in most games, could plausibly be physical or mental, power or precision, which is why it ends up ill-defined and negotiable. Frontiers has no such Attribute. Social force lives inside ◬ Resolve as the mental version of endurance, and nothing floats between quadrants.
The names¶
The names took a while to decide on, and there were many variations through development, but I landed on these because the goal was twofold: avoid the familiar terms, and make sure each word says plainly what it covers. Strength, Dexterity, and Intelligence come pre-loaded with what players expect them to mean, and each of the four covers more ground than those labels suggest.
- ❖ Vigor reads as constitution and natural strength together.
- ✩ Deft describes both fine precision, like sleight of hand, and gross precision, like footwork and reflexes.
- ◉ Acuity is clarity of mind combined with raw intelligence, wider than "smart."
- ◬ Resolve covers the social and mental wall a character holds, and reads as constitutional as well, which is why it can lend to durability.
Deft sits as an adjective among three nouns, and it was kept because it fit the capacity better than any noun that got tried.
Score and Rank¶
Every Attribute is read two ways: the Score when you're not rolling, the Rank when you are. In the roll-over Rulesets a Score defaults to 10 and the Rank is the Score minus 10: flat, linear, no tiers. The d100 runs the same structure on a wider scale, where the default is 50 and the Rank is (Score − 50) ÷ 5. The scale differs; the Rank it produces does not, which is why every value built on Rank is identical across all three.
The two-reading structure exists because passive resolution is a real part of the engine. Since a character can take their Score instead of rolling, the standing value needs to be a first-class number instead of something worked back from the modifier. Score is that number. Rank is the rolled value, and it's called Rank and not Modifier because Modifier only describes what it does to a roll, while Rank is more genre-agnostic and travels better across systems that might not roll the same way.
The conversion is linear rather than stepped, where a Rank changes every two points. (Figures below are given on the roll-over scale; the d100's equivalents follow from its own conversion.) Stepped scales invite an argument that never ends, over whether the odd or the even point is the one that counts, and they waste half the range. Linear means every point of Score is worth one point of Rank, and it scales evenly in both directions: a designer wanting a wider or narrower spread can stretch or compress the range, because there's no conversion to break.
Derivatives, and why Defense uses pairs¶
Most of the sheet is derived: ✚ HP, the ⊗ Wound and ⊖ Fatigue tracks, ⇉ Movement, ➤ Initiative, and both Defenses all come from Attributes. Derivation is what gives an Attribute its weight, since every value built on it is another reason to raise it.
One value sits alongside the Attributes instead of being derived from them: ϟ Energy, the resource that governs how far a character can reach past their level. It behaves like an Attribute-adjacent stat of its own, and it's documented in full in the Abilities documentation, which is where its reasoning lives.
Two derivation choices are worth the reasoning.
Defense draws on a pair¶
⛊ Physical Defense uses half the sum of ❖ Vigor and ✩ Deft; ⛉ Mental Defense uses half the sum of ◉ Acuity and ◬ Resolve. Two reasons. First, most systems give physical defense a standing value and route the mental side through a separate saving throw subsystem instead. Frontiers wanted mental defense to be a standing Defense equal to the physical one, so everything in the game resolves the same way: beat a number. No save-versus-attack split, and no second resolution path to learn. Second, pairing rewards rounded builds. A blow is turned by toughness and speed, an assault on the mind by sharpness and will, so a character leaning on a single stat defends worse than one balanced across the pair. Halving the sum keeps Defenses in a tight band, so no one becomes unhittable.
◬ Resolve lends to ✚ HP¶
❖ Vigor drives HP, but ◬ Resolve contributes as a character levels. It's the mental counterpart to bodily durability, the will to keep standing, and it gives ◬ Resolve a stake in survival rather than only in the mind.
Caps, and Live versus Anchored¶
The ⊗ Wound and ⊖ Fatigue tracks cap at 5 and do not grow with Level. Past a point, more boxes stop meaning anything and only slow play down, so the tracks stay short and readable for the whole life of a character. ✚ HP, which is meant to grow, is the pool that scales.
Derivatives are also marked Live or Anchored: Live values recalculate whenever their Attributes change, while Anchored values are fixed when set. The distinction lets designers write abilities that temporarily move an Attribute and have the right things follow, without the whole sheet shifting underfoot. It's an option instead of an obligation, but it's cheaper to offer now than to retrofit later.
Is this the best answer?¶
This is the part of the engine that feels most settled. The four-Attribute grid comes out complete without redundancy, every Attribute has work every character needs, and the Score-and-Rank structure leaves the numbers open to tune.
That last point is the important one for a designer. Almost everything here is a default rather than a fixture: the Score baseline, the linear conversion, the Defense formula, the caps. The structure is the opinion. The numbers are yours.
There's a good chance you want different numbers, and possibly a different count. The final section below sets out variant and alternative rules to use in their place, along with what each one changes and what it'll cost you elsewhere.
Variants¶
Variants keep the philosophy intact. Four Attributes, each read as a Score and a Rank, with the sheet derived from them. The numbers change while the structure holds, so derivation keeps working throughout. The groups below are dials, and they're meant to be mixed.
Tune the scale¶
The Score baseline is 10 and the Rank is the Score minus 10.
Shift the baseline¶
Move the default Score up or down. A lower baseline makes Ranks negative more often and gives the game a grittier, less-capable feel; a higher one does the reverse. Everything derived from Rank follows automatically.
For instance, a default Score of 8 puts a starting character at Rank −2 on anything they haven't invested in, so untrained attempts miss more often than they land.
Stretch or compress the range¶
Change what a point of Score is worth, for example a Rank of one per two points of Score, for a tighter spread where Attributes matter less against the dice. This is the one scale change that touches the linear conversion, so decide it early; every derived value inherits it.
For instance, one Rank per two points of Score turns a Score of 16 from Rank +6 into Rank +3, halving how hard an Attribute pushes against the dice.
Tune the derivation¶
The formulas hanging off the Attributes are all defaults.
Single-stat Defenses¶
Base each Defense on one Attribute rather than a pair. Simpler to calculate, and it makes specialists better defended, at the cost of the rounded-build reward the paired formula was built to give.
For instance, ⛊ Physical Defense reading straight off ❖ Vigor leaves the heavy bruiser harder to hit than the pair would allow, and the quick, lightly built character easier.
Un-cap the tracks¶
Let ⊗ Wounds and ⊖ Fatigue grow past 5, or scale with Level. Longer tracks make characters more durable and combat longer, and hand back the bookkeeping the cap removed.
For instance, letting ⊗ Wounds run to 8 at higher Levels gives a character three more to absorb before going down, and the GM three more boxes to track for every combatant on the field.
Re-weight ✚ HP¶
Change how much ❖ Vigor and ◬ Resolve contribute, or add a third contributor. The quickest lever on how survivable the game feels.
For instance, giving ◬ Resolve equal weight with ❖ Vigor leaves the scholar about as hard to kill as the soldier, which is a different game from the one where ❖ Vigor leads.
Tune what each stat drives¶
Re-map which Attribute feeds what. Move Skill count off ◉ Acuity, change what lends to ✚ HP, point a Defense at a different pair.
This is the riskiest group, because the dump stat problem lives here. Every derived value you move is a reason to invest taken from one Attribute and handed to another, and it's possible to re-map the sheet into a state where one Attribute is the obvious pick and one the obvious skip. If you change what a stat drives, check afterward that all four still have a job every character needs.
Alternatives¶
Alternatives replace the philosophy rather than tune it, and here that's expensive. The four-Attribute, Score-and-Rank structure is the single most-derived thing in the engine, so almost everything else (Defenses, ✚ HP, the tracks, Skill count) is built on top of it. Changing the count or the structure means rebuilding all of it.
One thing to hold in mind throughout: ϟ Energy is effectively a second Attribute-like system, a stat that governs how far a character can reach past their limits. Several of the alternatives below overlap with what Energy already does, so any restructuring of Attributes should account for it instead of treating it as separate. The reasoning behind Energy lives in the Abilities documentation.
More or fewer Attributes¶
The obvious swap. Go to six for finer distinctions, or three for a leaner sheet.
Six reintroduces the dump stat problem this design was built to avoid, and you'll need to find a real, universally-needed job for each of the six or accept that some will be ignored. Three loses a quadrant of the grid, so decide which axis collapses, usually by merging the two mental Attributes or the two physical ones. Every derived formula that names a specific Attribute has to be rewritten either way, and any that named a pair has to be reconsidered from scratch.
Attributes as spendable pools¶
Make each Attribute a resource a character burns, in the manner Cypher popularized: the stat is both how capable you are and a reserve you burn to push a roll or soak a hit.
This is compelling and complicated on Frontiers, because ϟ Energy already occupies much of that space. You'd either be building a second spendable system alongside Energy, which is a lot of overlapping bookkeeping, or folding Energy into the Attributes and rebuilding both at once. Worth attempting only if the spend-your-stats feel is central to your game, and worth reading the Energy documentation first to see what you'd be duplicating.
Minimal derivation¶
Keep four Attributes but derive almost nothing from them, after the fashion of Into the Odd. The Score is the number, and character identity moves onto Equipment and Abilities instead of a calculated sheet.
Cheaper than it looks, since removing derivation is mostly deletion. What it costs is the weight Attributes carry: with nothing built on them, raising an Attribute stops rippling outward, so advancement has to find its meaning elsewhere, usually in gear and abilities. Suits fast, lethal, improvisational play.
No Attributes at all¶
Remove the layer, as Fate does, and fold innate capability into Skills, Abilities, and the fiction.
This is the most expensive removal in the engine, the opposite pole from where Frontiers sits. Everything derived from an Attribute has to be re-rooted on something else, and the passive-resolution structure, taking a Score without rolling, has no standing number left to take. Realistically this isn't a modification of Frontiers so much as a different engine that borrows its other parts. Possible, but know going in that you're rebuilding the foundation.
Further afield¶
A few directions worth knowing about, each a larger departure, offered as references rather than worked options:
- Roll-based lifepath generation, where Attributes emerge from a character's history instead of being assigned, as in Traveller. This changes character creation more than the Attributes themselves, and it sits comfortably on the four-stat frame if you want origins to feel earned.
- Attributes that shift in play, rising and falling as a character is changed by events rather than only through advancement. Some horror and transformation-focused games do this, and it pairs naturally with the Live derivation the engine already supports.
- Asymmetric or archetype-specific stat sets, where different kinds of character use different Attributes entirely, as some tactical and mecha games do. This is the furthest departure here, and it breaks the shared-chassis assumption the whole engine rests on, so treat it as inspiration for a new system rather than a variant of this one.
Whichever you choose, remember that Attributes are the most-derived layer in Frontiers. A change here reaches further than a change anywhere else, so make it first, before building the rest of your system on top of it.